📚 A-Level Physics: Experimental Data Processing and Analysis Techniques | A-Level 物理：实验数据处理与分析技巧
Experimental data processing and analysis is a core skill in CIE A-Level Physics. Even with a perfect experimental setup, incorrect data handling can lead to wrong conclusions. This article provides a systematic guide to the key techniques you need to master for Paper 3 (Practical) and Paper 5 (Planning, Analysis and Evaluation).
实验数据处理与分析是 CIE A-Level 物理的核心技能。即使实验装置完美，错误的数据处理也会导致错误的结论。本文为你在 Paper 3（实验操作）和 Paper 5（规划、分析与评估）中需要掌握的关键技巧提供系统指导。
1. Precision, Accuracy, Uncertainty and Error | 精密度、准确度、不确定度与误差
Precision refers to the closeness of repeated measurements to each other, while accuracy refers to how close a measurement is to the true value. Uncertainty quantifies the range within which the true value is expected to lie. Error is the difference between the measured value and the true value — it can be systematic or random.
精密度指重复测量结果之间的接近程度，而准确度指测量值与真值的接近程度。不确定度量化了真值可能落在的范围。误差是测量值与真值之间的差异——可分为系统误差和随机误差。
Systematic error: arises from faulty calibration or flawed method; affects accuracy, not precision. It shifts all readings consistently in one direction.
系统误差：源于仪器校准不当或方法缺陷；影响准确度而非精密度。它使所有读数一致地偏向一个方向。
Random error: arises from unpredictable fluctuations in the environment or observer; affects precision. Repeated readings and averaging reduce its effect.
随机误差：源于环境或观察者的不可预测波动；影响精密度。重复读数并取平均可减小其影响。
Instrument uncertainty instruments, or ± the smallest division for digital instruments (unless stated otherwise).
仪器不确定度：对模拟仪器通常取最小分度的一半，对数字仪器取最小分度（除非另有说明）。
2. Representing Uncertainty: Absolute, Fractional and Percentage | 不确定度的表示：绝对、分数与百分比
An uncertainty can be written in three equivalent forms. Absolute uncertainty has the same unit as the measurement; fractional uncertainty is the ratio of absolute uncertainty to the measured value; percentage uncertainty is the fractional value multiplied by 100%.
不确定度可以用三种等价形式表示。绝对不确定度与测量值具有相同单位；分数不确定度是绝对不确定度与测量值的比值；百分比不确定度是分数值乘以 100%。
| Form | Expression | Example (R = 5.0 Ω ± 0.2 Ω) |
| Absolute 绝对 | x ± Δx | 5.0 ± 0.2 Ω |
| Fractional 分数 | Δx / x | 0.2 / 5.0 = 0.04 |
| Percentage 百分比 | (Δx / x) × 100% | 4% |
e measurement and its absolute uncertainty. For example, write 5.0 ± 0.2 Ω, not 5 ± 0.2 Ω.
记录原始数据时，测量值与绝对不确定度应保留相同的小数位数。例如，应写为 5.0 ± 0.2 Ω，而不是 5 ± 0.2 Ω。
3. Combining Uncertainties: Addition, Multiplication and Powers | 不确定度的合成：加、乘与幂
When combining measurements, uncertainties propagate differently depending on the mathematical operation. You must use the correct rule in every calculation.
当组合多个测量值时，不确定度的传播方式取决于数学运算类型。每次计算中必须使用正确的规则。
Addition/Subtraction (加/减): add absolute uncertainties. If Y = A + B or Y = A − B, then ΔY = ΔA + ΔB.
加/减：绝对不确定度相加。若 Y = A + B 或 Y = A − B，则 ΔY = ΔA + ΔB。
Multiplication/Division (乘/除) ctional (or percentage) uncertainties. If Y = A × B or Y = A / B, then ΔY/Y = ΔA/A + ΔB/B.
乘/除：分数（或百分比）不确定度相加。若 Y = A × B 或 Y = A / B，则 ΔY/Y = ΔA/A + ΔB/B。
Powers (幂) inty by the power. If Y = Aⁿ, then ΔY/Y = n × (ΔA/A). This also applies to roots, e.g. Y = √A = A1/2 gives ΔY/Y = (1/2)(ΔA/A).
幂：分数不确定度乘以幂指数。若 Y = Aⁿ，则 ΔY/Y = n × (ΔA/A)。此规则也适用于根号，例如 Y = √A = A1/2 时 ΔY/Y = (1/2)(ΔA/A)。
Constant multiples (常数倍): multiply the absolute uncertainty by the constant. If Y = kA, then ΔY = kΔA.
常数倍：绝对不确定度乘以该常数。若 Y = kA，则 ΔY = kΔA。
Example: T = 2π√l/g m, then ΔT/T = (1/2)(Δl/l) = 0.5 × 0.01 = 0.005. So T has a fractional uncertainty of 0.5%.
这意味着在计算百分不确定度时，系数 1/2 来自平方根。熟悉这些规则能让你在考试中快速且正确地处理不确定度传播。
This means the factor 1/2 comes from the square root. Mastering these rules allows you to propagate uncertainties quickly and correctly in exams.
4. Best-Fit Lines and Error Bars | 最佳拟合线与误差条
When plotting experimental data, each point that has an uncertainty should be plotted with an error bar showing the range of possible values. A best-fit line is a smooth straight line or smooth curve that best represents the trend of the data, balancing the points above and below it.
绘制实验数据时，每个具有不确定度的点都应画出误差条，显示可能的取值范围。最佳拟合线是一条能最好地代表数据趋势的光滑直线或曲线，使数据点均匀分布在线的上下两侧。
Key rules for a best-fit straight line (最佳拟合直线要点):
- Use a sharp pencil and a transparent ruler. 使用削尖的铅笔和透明直尺。
- Ensure roughly equal numbers of points lie above and below the line. 确保大致相等数目的点位于线上方和下方。
- Ignore obvious outliers when drawing the line, but plot them on the graph. 绘制直线时忽略明显异常点，但仍需在图上标出。
- Do not force the line through the origin unless theory demands it — test whether the intercept is consistent with zero. 除非理论要求，不要强行让直线过原点——检查截距是否与零一致。
- The line should extend across the full range of plotted data, not just the middle region. 直线应延伸至数据点的整个范围，而不仅是中间区域。
Error bars indicate the uncertainty in the dependent variable (y) and, where appropriate, the independent variable (x). If the uncertainty in x is negligible, draw only vertical error bars.
误差条表示因变量（y）的不确定度，在适当情况下也需表示自变量（x）的不确定度。如果 x 的不确定度可忽略，只需画垂直误差条。
5. Linearisation: Turning Non-Linear Relationships into Straight Lines | 线性化：将非线性关系转化为直线
Many physical relationships are non-linear, but a straight-line graph is easier to analyse and gives clearer evidence of the relationship. Linearisation involves rearranging the equation and choosing appropriate axes so that the plotted graph is a straight line. The gradient and intercept then correspond to physical quantities.
许多物理关系是非线性的，但直线图更易于分析，并能更清晰地验证关系。线性化涉及重排方程并选择合适的坐标轴，使绘制的图形为直线。斜率和截距则对应着物理量。
| Original relationship (原关系) | Plot (绘图) | Gradient (斜率) | Intercept (截距) |
| y = a x² | y against x² | a | 0 |
| T = 2π√l/g | T² against l | 4π²/g | 0 |
| V = E − Ir | V against I | −r | E |
| y = a ebx | ln y against x | b | ln a |
| y = a xⁿ | ln y against ln x | n | ln a |
For the exponential relationship y = a ebx nt b and intercept ln a. In all cases, check that the transformed axes produce a linear plot before calculating gradient and intercept.
对于指数关系 y = a ebx，取自然对数得到 ln y = ln a + bx。这是 y = mx + c 形式的直线方程，斜率为 b，截距为 ln a。在所有情况下，计算斜率和截距之前，都要先确认变换后的坐标轴能产生线性图。
6. Gradient and Intercept Calculations | 斜率和截距的计算
To calculate the gradient of a best-fit straight line, choose two points that lie on the line, not original data points unless they happen to lie exactly on the line. The points should be far apart to minimise percentage uncertainty. Read the coordinates of both points carefully, including their units.
计算最佳拟合直线的斜率时，应选择位于线上的两个点，而不是原始数据点（除非它们恰好落在线上）。两点应相距较远，以减小百分比不确定度。仔细读取两点的坐标（包括单位）。
m = (y₂ − y₁) / (x₂ − x₁)
Use the triangle method: draw a large right-angled triangle with the line as the hypotenuse, then divide the vertical difference by the horizontal difference. The intercept is read from the y-axis where the line crosses it. If the line is extrapolated, extend it carefully beyond the plotted region to read the intercept.
使用三角形法：以直线为斜边画一个大直角三角形，用垂直差除以水平差即为斜率。截距从直线与 y 轴的交点处读取。若需外推，小心地将直线延伸出数据区域以读取截距。
Uncertainty in the gradient can be estimated by drawing the worst acceptable line (steepest or shallowest) that still fits the error bars, then calculating the gradient of that line. The uncertainty is the difference between the best gradient and the worst gradient.
斜率的不确定度可通过绘制仍能拟合误差条的最陡或最缓的直线来估算，然后计算该直线的斜率。不确定度为最佳斜率与最差斜率之差。
7. Using logarithms for Non-Linear Data | 使用对数处理非线性数据
inst log x gives a straight line with gradient n and intercept log a. This is a powerful method because the exponent n can be read directly from the graph.
对于 y = a xⁿ 形式的关系，两边取对数得到 log y = n log x + log a。以 log y 对 log x 作图得到直线，斜率即为 n，截距为 log a。这是一种强大的方法，因为指数 n 可以直接从图中读出。
e intercept must be converted back using the inverse operation: if you used ln, then a = eintercept; if log₁₀, then a = 10intercept.
在 CIE 考试中，你可以使用常用对数（log₁₀）或自然对数（ln），但必须保持一致并说明使用哪一种。截距必须用逆运算转换回原始值：若使用 ln，则 a = e截距；若使用 log₁₀，则 a = 10截距。
When plotting log values, choose a sensible scale. Avoid using log paper if a calculator can provide log values, but if you do use log tables, keep at least 2 decimal places in your computed values.
绘制对数值时，选择合理的比例。如果计算器能提供对数值，就避免使用对数坐标纸；如果使用对数表，计算值至少保留 2 位小数。
8. Identifying and Handling Anomalous Data | 异常数据的识别与处理
An anomalous point is a data point that does not follow the trend of the majority of measurements. It may arise from a recording error, a sudden change in conditions, or a faulty reading. Anomalous points should not be removed automatically — you must first identify a possible cause.
异常点是不遵循大多数测量趋势的数据点。它可能源于记录错误、环境突变或仪器读数故障。不应自动删除异常点——必须先找到可能的原因。
Steps to handle anomalies (处理异常点的步骤):
- Plot the data as you go during the experiment, so anomalies are visible immediately. 实验过程中边做边画图，这样异常点立即可见。
- Check whether the anomaly was caused by a mistake in recording or calculation. If so, correct it. 检查异常是否由记录或计算错误引起。如果是，则予以更正。
- If the anomaly cannot be explained, repeat the measurement if time allows. 如果异常无法解释，在时间允许的情况下重复测量。
- When plotting the graph, plot the anomaly but exclude it from the best-fit line. Label it clearly. 绘图时标出异常点，但不计入最佳拟合线。明确标注该点。
- In your conclusion, mention the anomaly and its likely cause. 在结论中提及异常点及其可能原因。
In calculations using repeated measurements, you should discard only those readings that are clearly inconsistent with the spread of the others. Justification must be given in your report.
在使用重复测量的计算中，只能舍弃那些明显与其他读数分布不一致的数据。报告中必须给出理由。
9. Presenting Results: Tables, Graphs and Calculations | 结果呈现：表格、图表与计算
A well-presented results table is essential for gaining marks. Every column must have a clear heading with the physical quantity and its unit in the header, for example "current I / A" or "time t / s". Raw readings and derived quantities should be separated clearly.
设计良好的结果表格是得分的关键。每一列必须在表头标明物理量和单位，例如"电流 I / A"或"时间 t / s"。原始读数和导出量应清晰分开。
Column headers (列标题): write quantity name, symbol, slash, unit. Example: "length l / cm" or "l / cm". Do not repeat the unit in every cell.
列标题：写物理量名称、符号、斜杠、单位。例："长度 l / cm"或"l / cm"。不要在每一格中重复单位。
Significant figures (有效数字): raw readings should have the same number of decimal places, consistent with instrument resolution. Calculated values should not have more significant figures than the least precise input.
有效数字：原始读数应具有相同的小数位数，与仪器分辨率一致。计算值不应比最不精确的输入量有更多有效数字。
Graph axes (图轴): label both axes with quantity and unit. Choose a scale such that at least half of the graph paper is used in both directions. Use sensible intervals (1, 2, 5 or 10 units per square).
图轴：两个轴都标注物理量和单位。选择使图纸在两个方向上至少使用一半的比例。使用合理的间隔（每格 1、2、5 或 10 个单位）。
Calculations (计算): show your working. State the formula in symbols before substituting numbers. Keep units throughout.
计算：展示推导过程。先用符号写出公式，再代入数值。全程保留单位。
10. The Role of Significant Figures and Units | 有效数字与单位的作用
communicate the precision of a measurement. A reading of 2.50 cm is not the same as 2.5 cm — er about ±0.1 cm. Always record measurements with as many digits as the instrument allows.
有效数字传达测量的精密度。读数为 2.50 cm 与 2.5 cm 不同——前者暗示不确定度约为 ±0.01 cm，后者约为 ±0.1 cm。始终以仪器允许的位数记录测量值。
When performing calculations, the final answer should not have more significant figures than the measurement with the smallest number of significant figures used in the calculation. For example, if you measure length as 25.0 cm (3 s.f.) and time as 2.4 s (2 s.f.), the calculated speed should be given to 2 significant figures.
进行计算时，最终答案的有效数字不应超过计算中使用的最少有效数字的测量值。例如，若测得长度为 25.0 cm（3 位有效数字），时间为 2.4 s（2 位有效数字），则计算的速度应保留 2 位有效数字。
Units must be included in every numerical answer. Convert all values to SI base units where necessary before substitution. For example, convert g from grams to kilograms and cm to m in calculations, unless the question explicitly states otherwise.
每个数值答案都必须包含单位。必要时在代入计算前将所有值转换为 SI 基本单位。例如，计算中应将克转换为千克、厘米转换为米，除非题目另有明确说明。
11. Common Errors to Avoid in Practical Exams | 实验考试中应避免的常见错误
Many marks are lost in practical exams because of small but repeated mistakes. Being aware of these pitfalls will help you avoid them.
许多分数在实验考试中因为小而重复的错误被扣掉。了解这些陷阱将帮助你避免它们的发生。
| Mistake (错误) | Correction (纠正方法) |
| Drawing the line through all points including outliers 让直线穿过包括异常点在内的所有点 | Draw the best-fit line, ignoring clear outliers 绘制最佳拟合线，忽略明显异常点 |
| Forcing the line through the origin 强制让直线过原点 | Only do this if theory requires it and the intercept is consistent with zero 仅当理论要求且截距与零一致时才这样做 |
| Using data points rather than on-line points for gradient 用原始数据点而非线上的点求斜率 | Choose two points that lie exactly on the best-fit line 选择精确落在最佳拟合线上的两个点 |
| Incorrect scale choice — too small graph 比例选择不当——图太小 | Use at least half the graph paper; choose easy intervals 至少使用图纸的一半；选择易读的间隔 |
| Mixing decimal places in repeated readings 重复读数中小数位数不一致 | Keep the same number of decimal places for all readings of a quantity 同一物理量的所有读数保持相同的小数位数 |
| No error bars on graph points 图上的点没有误差条 | Add vertical error bars for y uncertainty; horizontal if x uncertainty is significant 为 y 不确定度添加垂直误差条；如果 x 不确定度显著则添加水平误差条 |
ncertainties to one significant figure, rounding the measured value to match: 5.0 ± 0.2 Ω.
另一个常见错误是不确定度的有效数字比测量值多。像 5.0 ± 0.23 Ω 这样的结果书写不当。不确定度一般保留一位有效数字，并将测量值舍入到一致精度：5.0 ± 0.2 Ω。
12. Final Checklist for Data Analysis | 数据分析最终检查清单
Before submitting your experimental report or answering any data-analysis question, run through this checklist to ensure you have not missed essential details.
在提交实验报告或回答任何数据分析问题之前，逐项检查以下清单以确保没有遗漏关键细节。
- All table columns have quantity, symbol and unit in the header. 所有表格列的表头都包含物理量、符号和单位。
- Repeated readings are shown and averaged where appropriate. 适当时显示重复读数并取其平均值。
- Graph axes are labelled with quantity and unit, with a suitable scale. 图轴标有物理量和单位，比例合适。
- All plotted points are visible and accurate to within half a small square. 所有绘图点清晰可见，准确度在半小格之内。
- Error bars are drawn correctly and matched by a worst-fit line where required. 误差条绘制正确，并在需要时配有最差拟合线。
- A best-fit straight line or smooth curve is drawn, not a dot-to-dot line. 绘制的是最佳拟合直线或光滑曲线，而不是逐点连线。
- Gradient calculation uses two points on the line, with coordinates clearly read. 斜率计算使用线上两点，坐标读数清晰。
- Intercept is read or calculated with a stated method (from the graph or from the linear equation). 截距的读取或计算有明确方法（从图中或从直线方程得出）。
- Uncertainty is propagated through all calculations using the correct rules. 不确定度使用正确规则传播到所有计算中。
- Final answers have correct units and appropriate significant figures. 最终答案具有正确单位和适当的有效数字。
Data analysis is not just about getting the number right — it is about demonstrating a logical, consistent and careful approach. Examiners award marks for method and clarity as much as for the final value.
数据分析不仅仅是得出正确的数字——更是展示一种逻辑、一致且严谨的方法。考官评分时，对方法和清晰度的重视程度与最终数值相同。
Published by TutorHao | Physics Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
或截图后用「扫一扫 → 相册」选择本图